Please join us for Ling-lunch this week:
Speaker: Coppe van Urk
Time: Thurs 12/3, 12:30-1:45
Place: 32-D461
Title: Movement and Antecedency in Obligatory Control
The two most prominent contemporary theories of control echo early theories in transformational grammar, Equi-NP Deletion (Rosenbaum 1967) and Postal’s (1970) null pronoun Doom. Hornstein (1999) suggests that control is derived through movement. Landau (2000 et seq.), on the other hand, argues for a PRO analysis. Regardless of what conceptual problems are associated with each analysis, there are significant empirical arguments for both approaches. In this talk, I go over some of these and conclude that both strategies are necessary. The empirical observation behind this is that some control complements behave like raising: the thematic positions have to be non-distinct, can share a single Case and the lower copy can be spelled out. Others behave more like non-obligatory control: they allow partial control and the lower position needs to have independent Case. These properties correlate across languages. I show that this account greatly simplifies the theory of control. Many of the special mechanisms that are necessary in Hornstein’s and Landau’s account can be dispensed with. In addition, the empirical coverage derived in this way is superior to that of other theories.